LinkR research information
Project funding from the Google Research Awards
This research project is funded by the Google Research Awards. We have no obligations to and will not provide Google with collected data or information about the project, except performance and other aggregate statistics that will be published. Please see our Privacy Policy agreement.
Abstract
Facebook users share links (URLs) with their friends passively on their own profile or actively by posting on their friends' profile. Everyday there are many links shared, but often Facebook users only look at links that match their interests or are 'popular' by some measure. Facebook provides simple methods of displaying a user's and his/her friends' shared links. Our research Facebook Application, Link Recommender (LinkR), recommends links to users based on the (personal) data they have shared with Facebook. Different recommending algorithms suggest the 'best' order of links to present to the user, based on the user's Facebook data. The 'best' order is loosely defined as those links the user is likely to click on and potentially share. Feedback from users through their usage or rating provides us with statistics to determine which algorithms are suitable for social networks.
